Iodometry is known as a method for determining the concentration of a halogen gas. The iodometry is a widely employed titration method for determining a halogen indirectly by oxidizing the iodine ion, a weak reducing agent, with a halogen, a strong oxidizing agent, to iodine, and titrating the liberated iodine. The iodometry is applicable to quantitative determination of a halogen such as chlorine and fluorine, and oxygen acid salts thereof.
A measurement apparatus employing the iodometry is disclosed in Japanese Patent Application Laid-Open No. 63-247655. This apparatus determines fluorine by converting fluorine to iodine gas which has a light absorption coefficient higher than the fluorine gas and has the maximum light absorption coefficient in the visible light region. This fluorine-detecting apparatus is provided with a converter for converting fluorine gas to iodine gas, and a detector for determining fluorine by optically detecting the iodine gas formed by conversion with the converter. Specifically, fluorine is converted to iodine gas in the fluorine/iodine converter. For example, the converter is constituted of a first reaction column packed with potassium chloride particles and a second reaction column packed with potassium iodide particles; fluorine gas is converted to chlorine in the first column packed with potassium chloride, and the chlorine is introduced into the second reaction column to react with the potassium iodide to liberate iodine (I2).
An indirect method for determining a halogen concentration is known in which a halogen gas is converted into another gas and measured the concentration of the gas and the concentration of the halogen gas is indirectly determined. For example, Japanese Patent Application Laid-Open No. 63-27736 discloses a gas concentration measurement apparatus which is constituted of a packing material for converting fluorine gas to another gas capable of absorbing infrared ray, a gas cell having an optical window transparent to the infrared ray on each of the side walls, a light source projecting infrared ray through the optical window, a detector for receiving the transmitted light from the gas cell, and a concentration meter for calculating the concentration of the converted gas from the output of the detector. The packing material is, for example, sulfur prepared by pulverizing crystalline sulfur. This packed sulfur material converts the fluorine gas to sulfur hexafluoride capable of absorbing infrared ray.
Any of the above methods measures the halogen concentration indirectly by converting the halogen gas once to another compound and measuring the concentration of the resulting converted compound. Therefore, the measurement results are obtained with a delay, which is not suitable for controlling the halogen gas concentration in a process at a prescribed level.
Usually, in a process for producing a halogen compound, the halogen gas is used desirably in excess of the substrate to conduct the halogenation effectively by keeping the halogen gas to be present in the reaction gas throughout the reaction. However, use of a large excess of the halogen gas is not economical owing to the necessity of a recovery system therefor. The amount of the halogen gas is decided depending on the kind of the substrate, the reaction temperature, and presence or absence of a catalyst. In a continuous production process, it is important to keep the halogen concentration within a suitable range. For measuring continuously the halogen concentration for example, the fluorine gas concentration, at a production site in the plant, methods are known as below:    (1) use of a fluorine detector employing an electrochemical cell;    (2) a method of introducing the objective gas into a potassium iodide solution and titrating automatically the iodine formed by reaction with fluorine gas by sodium thiosulfate; and so forth.
The above method (1), although it is useful for measuring temporarily fluorine in a gas containing the fluorine at a low concentration, is not useful for the object of the present invention, since the cell will deteriorate in a short time in continuous analysis of a fluorine containing gas. The above method (2), which automates a general method of fluorine determination in a gas, has disadvantages of low response speed owing to relatively long time required for the analysis, complicate maintenance, a relatively large space for the apparatus installation, and so forth.
Japanese Patent Application Laid-Open No. 2000-22255 describes a method of continuous analysis of fluorine in a gas. This method is applicable to direct measurement of a fluorine gas concentration in a gas mixture to control stably the fluorine gas concentration in real time. This method is employed for measurement of a concentration of a fluorine-containing gas mixture, for example, a gas used in an excimer laser apparatus by detecting the fluorine gas concentration from change in UV absorption of the fluorine gas. In this patent specification, a mixed gas is employed which contains Kr gas, or Ne gas in addition to the fluorine gas at a fluorine gas concentration of 1.0%, or 9.0%, for example.
However, this method involves problems in measurement of a fluorine gas concentration in a fluorine compound production process. Specifically, in a process for fluorinating a hydrocarbon or a hydrofluorocarbon by fluorine gas to produce a perfluorocarbon, the reaction generates a great amount of reaction heat: a larger amount of heat with increase of the amount of the fluorine, in proportion to the moles of the fluorine reacted. The great heat generation is liable to cause scission of the C—C bond, explosion, or the like, and lowering of the product yield, which are problems in industrial production and operation. To suppress the violent generation of the reaction heat, the fluorine gas is diluted with an inert gas (nitrogen, helium, etc.). Further, the method has another problem. That is, the gas after the reaction contains the perfluorocarbon as the reaction product and hydrogen fluoride equimolar to the substituted hydrogen, which lowers significantly the fluorine concentration. The continuous measurement of light absorbance of this fluorine in the gas mixture is difficult owing to the low fluorine concentration and influence of the other mixed gas depending on the maximum absorption wavelengths (λmax/nm) and the molar absorption coefficients (unit: mol−1dm3cm−1) of the component gases.
On the other hand, perfluorocarbons such as tetrafluoromethane and hexafluoroethane are used as an etching gas or a cleaning gas in semiconductor device production processes. Regarding the production process for such a fluorocarbon, various methods are known, for example, as shown below:    (1) reaction of ethane with F2 in a jet reactor to form tetrafluoromethane (hereinafter occasionally referred to as “FC-14” or “CF4”), or hexafluoroethane (hereinafter occasionally referred to as “FC-116” or “CF3CF3”) with nitrogen gas used as the diluent gas (J.Am.Chem.Soc., 77, 3007 (1955); J.Am.Chem.Soc., 82, 5827 (1960));    (2) fluorination of C—H with F2 in a reactor having a porous alumina tube (EP31519);    (3) fluorination of a linear hydrocarbon with F2 in the presence of a diluent gas in a reactor having a porous metal tube (double tube structure): SF6, CF4, C2F6 or C3F8 being used as the diluent gas (EP32210); and    (4) reaction of a saturated or unsaturated hydrocarbon, or a partially fluorinated hydrocarbon with F2 to produce a hydrofluorocarbon (U.S. Pat. No. 5,406,008), or reaction of an alkene with carbon containing F2 by adsorption to produce a fluorinated alkene (Japanese Patent Application Laid-Open No. 2-207052).
However, these methods are not necessarily satisfactory for the purpose of safe and efficient production of a perfluorocarbon.
Other methods are disclosed as below:    (5) reaction of a hydrofluorocarbon with F2 in a gas phase in the presence of a diluent gas at an elevated temperature (Japanese Patent Application Laid-Open Nos. 9-241186, and 9-241187). In these method, a perfluorocarbon is produced safely and efficiently by reaction with F2 at an elevated temperature by keeping the hydrofluorocarbon concentration at the reactor inlet to be not higher than 8 mole % by use of a diluent gas.
For conducting the direct fluorination reaction safely, the hydrofluorocarbon concentration at the reactor inlet should be analyzed and controlled precisely and quickly. Conventionally, in quantitative determination of a hydrofluorocarbon in a gas stream at the reactor inlet, a part of the mixed gas is sampled, and washed with a potassium iodide solution, and then (1) the acid gas components such as HF and F2 are determined quantitatively by iodometric titration or neutralization titration of a part of the washing solution, and (2) the washed gas is analyzed by gas chromatography to determine the hydrofluorocarbon, the perfluorocarbon, and other minor gas components.
However, this analysis method involves a certain time lag between the start of the sampling of the gas mixture and the completion of the analysis, and is not suitable for the reaction process control. Furthermore, the gas sampling, the iodometric titration, and the neutralization titration have to be conducted manually, which compels the worker to handle the sample materials containing dangerous corrosive HF and F2 gases.